EasyFL_with_KDE/applications/fedureid/evaluate.py

import numpy as np
import torch
import torch.nn as nn
from torch.autograd import Variable
import scipy.io


def extract_feature(model, dataloaders, device, ms=[1]):
    features = torch.FloatTensor()
    model = model.to(device)
    for data in dataloaders:
        img, label = data
        n, c, h, w = img.size()
        ff = torch.FloatTensor(n, 2048).zero_().to(device)
        for i in range(2):
            if i == 1:
                img = fliplr(img)
            input_img = Variable(img.to(device))
            for scale in ms:
                if scale != 1:
                    # bicubic is only  available in pytorch>= 1.1
                    input_img = nn.functional.interpolate(input_img, scale_factor=scale, mode='bicubic',
                                                          align_corners=False)
                _, outputs = model(input_img)
                ff += outputs
        # # norm feature
        fnorm = torch.norm(ff, p=2, dim=1, keepdim=True)
        ff = ff.div(fnorm.expand_as(ff))
        features = torch.cat((features, ff.data.cpu()), 0)
    return features


def test_evaluate(path, dataset, device):
    result = scipy.io.loadmat(path)
    query_feature = torch.FloatTensor(result['query_f'])
    query_cam = result['query_cam'][0]
    query_label = result['query_label'][0]
    gallery_feature = torch.FloatTensor(result['gallery_f'])
    gallery_cam = result['gallery_cam'][0]
    gallery_label = result['gallery_label'][0]
    query_feature = query_feature.to(device)
    gallery_feature = gallery_feature.to(device)
    CMC = torch.IntTensor(len(gallery_label)).zero_()
    ap = 0.0
    for i in range(len(query_label)):
        ap_tmp, CMC_tmp = evaluate(query_feature[i], query_label[i], query_cam[i], gallery_feature, gallery_label,
                                   gallery_cam)
        if CMC_tmp[0] == -1:
            continue
        CMC = CMC + CMC_tmp
        ap += ap_tmp
    CMC = CMC.float()
    CMC = CMC / len(query_label)  # average CMC
    return CMC[0], CMC[4], CMC[9], ap / len(query_label)


def evaluate(qf, ql, qc, gf, gl, gc):
    query = qf.view(-1, 1)
    score = torch.mm(gf, query)
    score = score.squeeze(1).cpu()
    score = score.numpy()
    # predict index
    index = np.argsort(score)  # from small to large
    index = index[::-1]
    # good index
    query_index = np.argwhere(gl == ql)
    camera_index = np.argwhere(gc == qc)

    good_index = np.setdiff1d(query_index, camera_index, assume_unique=True)
    junk_index1 = np.argwhere(gl == -1)
    junk_index2 = np.intersect1d(query_index, camera_index)
    junk_index = np.append(junk_index2, junk_index1)  # .flatten())

    CMC_tmp = compute_mAP(index, good_index, junk_index)
    return CMC_tmp


def compute_mAP(index, good_index, junk_index):
    ap = 0
    cmc = torch.IntTensor(len(index)).zero_()
    if good_index.size == 0:  # if empty
        cmc[0] = -1
        return ap, cmc

    # remove junk_index
    mask = np.in1d(index, junk_index, invert=True)
    index = index[mask]

    # find good_index index
    ngood = len(good_index)
    mask = np.in1d(index, good_index)
    rows_good = np.argwhere(mask == True)
    rows_good = rows_good.flatten()

    cmc[rows_good[0]:] = 1
    for i in range(ngood):
        d_recall = 1.0 / ngood
        precision = (i + 1) * 1.0 / (rows_good[i] + 1)
        if rows_good[i] != 0:
            old_precision = i * 1.0 / rows_good[i]
        else:
            old_precision = 1.0
        ap = ap + d_recall * (old_precision + precision) / 2

    return ap, cmc


def fliplr(img):
    inv_idx = torch.arange(img.size(3) - 1, -1, -1).long()  # N x C x H x W
    img_flip = img.index_select(3, inv_idx)
    return img_flip